Linguistic analysis of seed documents and peer groups

Abstract

Systems may perform analyses of claims included in a patent document. The systems may generate one or more search strings from the patent document and provide the one or more search strings to a third-party searching authority. The third-party searching authority may return a collection of documents responsive to the one or more search strings. In particular situations, the systems may re-rank the documents of the collection to provide a patent centric ranking. The systems may also analyze the documents of the collection with respect to the elements of the claims to generate various types of patent infringement and/or invalidity reports.

What is claimed is: 
     
       1. A method comprising:
 receiving a seed document including content; 
 determining that the seed document is a patent document and that the content includes a patent claim, wherein the patent document is classified according to a classification; 
 generating, based upon a natural language processing technique, noun phrases for the patent claim within the seed document, wherein the natural language processing technique segments the patent claim, performs syntactic parsing, generates tagging data associated with the patent claim, and utilizes a claim-specific grammar model to generate the noun phrases; 
 generating a trained machine learning model configured to identify elements of patent claims, the trained machine learning model trained on a document corpus specific to the seed document; 
 analyzing the noun phrases to identify an element of the patent claim, the element comprising a plurality of syntactically related concepts in the patent claim, the element being one of a plurality of elements of the patent claim, and the element including a plurality of words, wherein the analyzing is performed utilizing the trained machine learning model; 
 determining a breadth of the element with respect to one or more additional elements of the plurality of elements of the patent claim; 
 generating, based at least partly on the breadth of the element, a search string that includes one or more of the words of the plurality of words of the element; 
 sending the search string to a third-party searching authority; 
 obtaining a plurality of additional patent documents, the plurality of additional patent documents being classified according to the classification and the plurality of additional patent documents including a first set of additional patent documents having a first priority date before a priority date of the patent document and a second set of additional patent documents having a second priority date after the priority date of the patent document; 
 determining a first frequency of occurrence of a word of the element of the claim in the first set of additional patent documents; 
 determining a second frequency of occurrence of the word of the element of the claim in the second set of additional patent documents; 
 analyzing the first frequency of occurrence of the word with respect to a first threshold frequency of occurrence; and 
 analyzing the second frequency of occurrence of the word with respect to a second threshold frequency of occurrence, the second threshold frequency of occurrence being greater than the first threshold frequency of occurrence. 
 
     
     
       2. The method as recited in  claim 1 , further comprising:
 performing an analysis of words included in individual documents of the plurality of additional patent documents with respect to the plurality of words included in the element; 
 determining, based at least partly on the analysis, a portion of a document included in the plurality of additional patent documents that includes at least a portion of the plurality of words included in the element; 
 generating a mapping indicating an association between the claim element and the portion of the document; and 
 generating data of a user interface that includes the mapping. 
 
     
     
       3. The method as recited in  claim 2 , wherein:
 the user interface includes a claim chart including at least a first column, a second column, and at least one row; 
 the first column includes the claim element and the second column includes the portion of the document; and 
 the claim element and the portion of the document are disposed in a same row of the claim chart. 
 
     
     
       4. The method as recited in  claim 1 , further comprising:
 performing an analysis of words included in portions of the seed document other than a claims portion of the seed document with respect to the plurality of words included in the element; 
 determining, based at least partly on the analysis, a portion of the seed document that includes at least a portion of the plurality of words included in the element; 
 generating a mapping indicating an association between the claim element and the portion of the seed document; and 
 generating data of a user interface that includes the mapping. 
 
     
     
       5. The method as recited in  claim 4 , further comprising:
 determining a measure of validity of the claim based at least partly on the mapping; and 
 generating a user interface that includes the measure of validity. 
 
     
     
       6. The method as recited in  claim 1 , further comprising:
 determining that the patent claim is an independent claim; 
 identifying a dependent claim from the content of the seed document; 
 determining a frequency of occurrence of one or more terms included in the dependent claim in individual documents of the plurality of additional patent documents; and 
 ranking the plurality of additional patent documents based at least in part on the frequency of occurrence of the one or more terms of the dependent claim within the individual documents. 
 
     
     
       7. The method as recited in  claim 1 , further comprising:
 ranking the plurality of additional patent documents based at least in part on the first frequency of occurrence and the second frequency of occurrence. 
 
     
     
       8. The method as recited in  claim 1 , further comprising ranking the plurality of additional patent documents according to a set of criteria that is different from one or more criteria used by the third-party searching authority to rank the plurality of additional patent documents. 
     
     
       9. A system comprising:
 one or more communication interfaces; 
 one or more processors; and 
 non-transitory computer-readable media storing computer-executable instructions, that when executed by the one or more processors cause the one or more processors to perform operation comprising:
 obtaining a patent document that includes a claim, wherein the patent document is classified according to a classification; 
 generating, based upon a natural language processing technique, noun phrases for the patent claim within the patent document, wherein the natural language processing technique segments the patent claim, performs syntactic parsing, generates tagging data associated with the patent claim, and utilizes a claim-specific grammar model to generate the noun phrases; 
 generating a trained machine learning model configured to identify elements of claims, the trained machine learning model trained on a document corpus specific to the patent document; 
 analyzing, utilizing the trained machine learning model, the noun phrases to identify an element of the claim, the element comprising a plurality of syntactically related concepts in the claim; 
 generating, based at least in part on a breadth value of the element, a search string that includes at least one word from the element of the claim; 
 sending, via the one or more communication interfaces, the search string to a third-party searching authority; 
 obtaining a plurality of additional patent documents, the plurality of additional patent documents being classified according to the classification and the plurality of additional patent documents including a first set of additional patent documents having a first priority date before a priority date of the patent document and a second set of additional patent documents having a second priority date after the priority date of the patent document; 
 determining a first frequency of occurrence of a word of the element of the claim in the first set of additional patent documents; 
 determining a second frequency of occurrence of the word of the element of the claim in the second set of additional patent documents; 
 analyzing the first frequency of occurrence of the word with respect to a first threshold frequency of occurrence; and 
 analyzing the second frequency of occurrence of the word with respect to a second threshold frequency of occurrence, the second threshold frequency of occurrence being greater than the first threshold frequency of occurrence. 
 
 
     
     
       10. The system as recited in  claim 9 , wherein:
 the non-transitory computer-readable media store further computer-executable instructions, that when executed by the one or more processors cause the one or more processors to perform further operations comprising:
 determining that the first frequency of occurrence of the word is less than the first threshold frequency of occurrence; and 
 determining that the second frequency of occurrence of the word is greater than the second threshold frequency of occurrence; and 
 wherein generating the search string includes adding the word to the search string based at least partly on the first frequency of occurrence of the word being less than the first threshold frequency of occurrence and the second frequency of occurrence of the word being greater than the second threshold frequency of occurrence. 
 
 
     
     
       11. The system as recited in  claim 9 , wherein the non-transitory computer-readable media store further computer-executable instructions, that when executed by the one or more processors cause the one or more processors to perform further operations comprising determining the breadth value of the element based at least partly on first frequency of occurrence of the word and the second frequency of occurrence of the word. 
     
     
       12. The system as recited in  claim 9 , wherein the non-transitory computer-readable media store additional computer-executable instructions, that when executed by the one or more processors cause the one or more processors to perform additional operations comprising:
 receiving a collection of documents from the third-party searching authority, the collection of documents being responsive to the search string and the collection of documents being ranked according to a first order; 
 modifying rankings of the collection of documents to generate a re-ranked collection of documents, the re-ranked collection of documents being ranked according to a second order; and 
 generating data of a user interface that includes at least a portion of the re-ranked collection of documents. 
 
     
     
       13. The system as recited in  claim 12 , wherein the non-transitory computer-readable media store further computer-executable instructions, that when executed by the one or more processors cause the one or more processors to perform further operations comprising:
 modifying the search string to generate a modified search string, the modified search string including an additional word that is different from one or more words included in the search string; 
 sending, via the one or more communication interfaces, the modified search string to the third-party searching authority; and 
 receiving an additional collection of documents from the third-party searching authority, the additional collection of documents being responsive to the additional search string. 
 
     
     
       14. The system as recited in  claim 13 , wherein:
 the non-transitory computer-readable media store other computer-executable instructions, that when executed by the one or more processors cause the one or more processors to perform other operations comprising determining an additional word included in at least one document of the plurality of additional patent documents or a portion of the patent document other than a claims portion of the patent document, wherein the additional word is semantically similar to the at least one word from the element of the claim included in the search string, and the modified search string includes the additional word. 
 
     
     
       15. A method comprising:
 obtaining a patent document that includes a claim, wherein the patent document is classified according to a classification; 
 generating, based upon a natural language processing technique, noun phrases for the patent claim within the patent document, wherein the natural language processing technique segments the patent claim, performs syntactic parsing, generates tagging data associated with the patent claim, and utilizes a claim-specific grammar model to generate the noun phrases; 
 generating a trained machine learning model configured to identify elements of claims, the trained machine learning model trained on a document corpus specific to the patent document; 
 analyzing the noun phrases to identify an element of the claim, the element comprising a plurality of syntactically related concepts in the claim, wherein the analyzing is performed utilizing the trained machine learning model; 
 generating, based at least in part on a breadth value associated with the element, a search string that includes at least one word from the element of the claim; 
 sending, via one or more communication interfaces, the search string to a third-party searching authority; 
 obtaining a plurality of additional patent documents, the plurality of additional patent documents being classified according to the classification and the plurality of additional patent documents including a first set of additional patent documents having a first priority date before a priority date of the patent document and a second set of additional patent documents having a second priority date after the priority date of the patent document; 
 determining a first frequency of occurrence of a word of the element of the claim in the first set of additional patent documents; 
 determining a second frequency of occurrence of the word of the element of the claim in the second set of additional patent documents; 
 analyzing the first frequency of occurrence of the word with respect to a first threshold frequency of occurrence; and 
 analyzing the second frequency of occurrence of the word with respect to a second threshold frequency of occurrence, the second threshold frequency of occurrence being greater than the first threshold frequency of occurrence. 
 
     
     
       16. The method as recited in  claim 15 , further comprising:
 determining a part of speech for individual words included in the claim; and 
 determining limitations included in the claim by identifying at least one of nouns, noun and adjective couplings, verbs, or verb and adverb couplings with respect to the words included in the claim; and 
 wherein the element of the claim includes a plurality of the limitations included in the claim. 
 
     
     
       17. The method as recited in  claim 16 , further comprising:
 determining a plurality of predicates included in the claim and a plurality of arguments included in the claim; 
 determining at least one predicate of the plurality of predicates that corresponds to an argument of the plurality of arguments; and 
 wherein the element of the claim includes at least a grouping of an argument of the plurality of arguments with a predicate of the plurality of predicates. 
 
     
     
       18. The method as recited in  claim 15 , wherein:
 the plurality of additional patent documents is ranked according to a first order; and 
 the method further comprises modifying rankings of the plurality of additional patent documents to generate a re-ranked collection of documents, the re-ranked collection of documents being ranked according to a second order. 
 
     
     
       19. The method as recited in  claim 18 , further comprising:
 determining a portion of a document included in the re-ranked collection of documents that includes at least a portion of words included in the element; 
 generating a mapping indicating an association between the claim element and the portion of the document; and 
 generating data of a user interface that includes a claim chart including at least a first column, a second column, and at least one row; and 
 wherein the first column includes the claim element, the second column includes the portion of the document; and the claim element and the portion of the document are disposed in a same row of the claim chart.